The silky weapon of Rollup? Understand the shared sorter Astria in one article

Author: 1912212.eth, Foresight News

With the increasingly mature Ethereum second-layer network and the significant drop in Gas fees, these networks are showing a booming rise trend. However, accompanying problems have emerged, such as fees and speed being affected by various factors and the user experience being unsatisfactory. Against this background, the importance of transaction sequencing is becoming increasingly prominent, becoming the key to solving transaction bottlenecks and optimizing user experience.

In April this year, Astria, a modular blockchain focusing on shared sorters, completed a $5.5 million seed round of financing, led by Maven 11, with participation from 1kx, Delphi Ventures, Robot Ventures, and others. Just three months later, Astria completed another $12.5 million financing round, led by dba and Placeholder VC, with participation from Hasu and others.

What is Astria?

Astria is developing a decentralized shared sequencer network aimed at providing rollups with fast finality, censorship resistance, composability, and decentralization.

Currently, L2 operates a centralized sequencer that is more convenient, cheaper, and easier for users to use, so most mainstream L2s are managed by their own teams. Although L2 users can directly submit transactions to L1 to bypass the sequencer, they must pay transaction gas fees to L1, and transactions may take longer to complete.

The sorter controls the order of transactions, theoretically it has the right not to include user transactions. The sorter can also extract MEV from the transaction pool. If there is only one sorter, the centralization risk will increase.

Therefore, a Decentralization shared sorter still has its meaning.

How Astria Works

The Astria Decentralization sorter has long sorters Node, so it can sort Rollup transactions. In Astria’s operation mode, users submit transactions to Rollups, and the transactions automatically enter their respective Rollup Node memory pools, where the combiners are responsible for collecting txns and sending them to the sorter. Finally, the sorter consolidates the txns into a shared block and sends a pre-confirmation to the user.

Current sorters are all based on specific Rollups. Astria processes Blocks in batches for long Rollups. Through data compression, more long costs can be saved when publishing data to L1. The shared sorter network of Decentralization will incentivize participants from long Rollup ecosystems to act as validators on the network.

Astria Stack

Astria’s main components include five major parts: Combiner, Sorting Layer, Relay, DA, and Scheduler.

Combiner

Technical professionals may be able to directly use the sorting layer for better trade sorting, but for the vast majority of ordinary users, it will increase the difficulty in practical use. Direct interaction with the sorting layer requires users to hold the Sorter Token and maintain the Sorter Wallet, both of which have a significant negative impact on user experience.

Astria provides users with combiners to abstract this complexity. Combiners are like a gas station, bearing the sorting cost of user transactions. Combiners also provide users with an unordered guarantee, bundling transactions in the order they are received.

Sorting layer

Astria’s sorting layer uses CometBFT as its ConsensusAlgorithm. Chains that support CometBFT can support IBC (Inter-Blockchain Communication), which means it can perform Cross-Chain Interaction among many other chains.

The uniqueness of the Astria Sorter lies in the fact that the transactions it contains are not executed (sorted by latency), but are assigned to another execution engine, Rollup. The Sorter Node can choose to act as ‘validators’, which means it actively participates in block production and final confirmation.

The application logic of the Astria sorter allows three main functions:

• Sorting of Rollup data
• Value Transfer
• Changes to Validator Set

Relay

The function of the relay is to obtain validated blocks from the sorter and pass them to the DA layer. Since the block time of the sorter is faster than the DA, the relay will first batch process the ordered data from long sorter blocks, then compress it, and submit it to the DA.

Each sorter block can also be obtained by the scheduler from the sorter before submitting it to the DA in the relay. This enables a quick final determination of the improved user experience, serving as a soft submission of the execution layer. The data set sent from the relay to the DA layer serves as a factual source and is ultimately extracted from the DA for a definitive confirmation submission in the Rollup.

DA

Astria uses Celestia as the data availability layer, which is the ultimate destination for all the data sorted in the sorter network. Once the data is written to Celestia, the transaction order is considered final, and all the data will be extracted from here when a new Rollup Node is started.

Scheduler

The scheduler can be thought of as the Consensus implementation of the Rollup Full Node, similar to the operation Node in the OP Stack. The scheduler is the corresponding part of the execution engine, together they form a complete Rollup Node. Its function is to extract transactions belonging to the Rollup Node from each sorter block and forward them to the execution layer, connecting the sorter and DA layers to the Rollup execution layer.

For each sorter block, extract the required Rollup data, verify the batch Rollup data, wait for verification to complete, and then convert it into a transaction list and pass it to the execution engine.